Shipping a sensitive change
Focus on new permissions, customer administration, billing, integrations, or identity flows and the existing features they affect.
One product, many customers, and plenty of places for permissions to go wrong. We test tenant isolation, user roles, SSO, and the workflows that connect your web app and APIs.
A new SSO feature needs a different look from a first assessment of the whole product. Start with the decision your team needs to make.
Focus on new permissions, customer administration, billing, integrations, or identity flows and the existing features they affect.
Check what they expect from the scope, report, and re-test before booking. Their reviewer decides whether the evidence meets the request.
Assess the connected product and give engineering a list of confirmed issues, priorities, and areas that still need a closer look.
An export, background job, or admin action still needs to respect the right customer's boundaries. We follow the workflows through those less visible parts.
Check records, files, search, exports, shared links, and background jobs using controlled tenants. Where sharing is allowed, we document the exception and test its limits.
Owners, admins, members, and guests need different permissions. We check what happens when someone is invited, moved, given ownership, or removed.
Review how SSO, account linking, recovery, sessions, and provisioning affect membership. We agree on the identity-provider configurations and test accounts first.
Check API keys, webhooks, service accounts, billing permissions, approvals, and background work. Third-party platforms stay out of scope without their authorization.
A short walkthrough of the modules, APIs, roles, and integrations helps us scope the work. For isolation checks, we need two controlled tenants and representative accounts.
Use a stable build, synthetic records, working role-specific accounts, API documentation, and a technical contact. Note differences from production, especially SSO, billing, feature flags, and background processing.
The SaaS pentest readiness checklist helps your team prepare these details.
Written authorization defines hosts, roles, safe data, request limits, contacts, and stop conditions. Production testing requires explicit safeguards. Destructive testing, load testing, cloud configuration review, and source review are not included by default.
Need broader coverage? See the full security capability map.
A tester verifies every reported finding. You'll see the evidence, the demonstrated impact, and what we couldn't check. AI assists the work; it doesn't decide whether an issue is confirmed.
The redacted draft-authorization story shows how we explain those distinctions. It's an educational reconstruction, not a client endorsement or confirmation of a deployed fix.
Agree on these details before reserving engineering time.
Roles, tenant relationships, workflow depth, integrations, and environment readiness drive effort more than screen counts. Reporting and re-test requirements also matter. We scope the work before quoting; the pentest pricing guide explains what to compare.
Usually they share identity and customer data, so connected coverage can be useful. The agreement names what is included. Compare web application testing and API testing when deciding where the boundary belongs.
It can provide scoped testing evidence, findings, and agreed re-test results. Confirm the recipient's requirements first. A penetration test is not a compliance certification, a guarantee of customer acceptance, or proof that the whole product is vulnerability-free.
Share your deadline and planned changes during scoping. We confirm availability after reviewing scope and access. Allow time for reporting, remediation, and any included re-test; new features or material changes may need a revised scope.
A short description, the main roles, and any deadline are enough to start. Leave credentials, tokens, and customer data out of the enquiry.